Volume 86, Issue 4, Pages 1055-1066 (May 2015) Inhibitory Neuron Transplantation into Adult Visual Cortex Creates a New Critical Period that Rescues Impaired Vision  Melissa F. Davis, Dario X. Figueroa Velez, Roblen P. Guevarra, Michael C. Yang, Mariyam Habeeb, Mathew C. Carathedathu, Sunil P. Gandhi  Neuron  Volume 86, Issue 4, Pages 1055-1066 (May 2015) DOI: 10.1016/j.neuron.2015.03.062 Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 1 Transplanted Cells Migrate in Adult Visual Cortex and Express Markers of Mature Cortical Interneurons (A) Schematic of physiologically guided transplantation to binocular visual cortex. (Right) A retinotopic map overlaid on an image of the cortical surface. Location of MGE cell injections and GCaMP6s-expressing virus injections indicated in red and green, respectively. (B) Example coronal section from a MGE recipient 95 days after transplantation (DAT). Transplanted cells disperse across cortical layers and express the GABAergic neuron marker VGAT (red). (C) Example transplanted VGAT-positive cells (left column) from a recipient ∼100 DAT stained for Parvalbumin (PV, top row), Somatostatin (SOM, middle row), and vasoactive intestinal peptide (VIP, bottom row). White chevrons show transplanted cells expressing PV or SOM. (D) Quantification of transplanted marker expression for PV, SOM, and VIP (n = 480 cells, n = 2 mice). (E) Example perineuronal nets (PNNs) on transplanted cells ∼100 DAT. White chevrons show transplanted cells that carry PNNs. Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 2 Transplanted Inhibitory Neurons Develop Cell-Type-Appropriate Visual Responses (A) A transplanted PV+ (red) cell coexpressing calcium indicator GCaMP6s (green). (B) GCaMP6s visual responses from a PV− cell (top, green) and transplanted PV+ cell (bottom, red) to drifting gratings presented at 12 different orientations (gray bars denote stimulus presentation, gray traces reflect single trial responses, red or green trace represents averaged signal). (C) Polar plots of averaged response to stimulus orientation are shown for each example trace. (D) Orientation selectivity of transplanted PV+ neurons (∼90 DAT; tPV+, solid red bar), endogenous PV+ neurons (∼P140; ePV+, striped red bar), and neighboring PV− neurons (tPV−, solid green bar; ePV−, striped green bar). Transplanted PV+ neurons have broader orientation tuning than their neighbors (OSI tPV+ = 0.39 ± 0.07, n = 11 versus tPV− = 0.72 ± 0.04, n = 17). This orientation tuning is equivalent to endogenous PV+ cells(OSI ePV+ = 0.46 ± 0.03, n = 19; p = 0.78). (E) Average response at preferred orientation (ΔF/F0) is shown for transplanted and endogenous PV+ (tPV+, solid red bar; ePV+, striped red bar) and PV− neurons (black bar). Responses from transplanted PV+ cells (ΔF/F0 = 0.42 ± 0.06) were comparable in strength to endogenous PV+ cells (ΔF/F0 = 0.38 ± 0.03). Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 3 Transplantation Reactivates Critical Period Plasticity (A) Timeline of the experimental protocol. Responses to contralateral versus ipsilateral eye stimulation were used to calculate an ocular dominance index (ODI). Recordings were made only in the transplanted hemisphere. (B) Ocular dominance index determined before (white columns) and after (gray columns) 4 days of monocular deprivation (4d MD). A significant difference was seen between pre- and post-MD ODI in the CP group (yellow, n = 6: W[5] = −21; p = 0.03) and in the 35 DAT MGE group (magenta, n = 9: W[8] = −45; p = 0.004), but not in any other group. (C) Average ocular dominance shift (ODS) following 4d MD for each experimental group in (B). ODS for MGE 35 DAT (magenta, n = 9) recipients was equivalent to critical period shifts (yellow, n = 6), but ODS in all other groups were significantly smaller than critical period ODS (versus untreated, gray, n = 7, p = 0.006; versus MGE 70 DAT, cyan, n = 8, p = 0.01; versus dead MGE, dashed magenta, n = 4, p = 0.04; versus LGE, black solid, n = 5, p = 0.03; and versus CGE, black crosshatched, n = 5, p = 0.02). (D) (Upper graph) Four days of MD produced a loss of deprived eye visual responses (n = 8; black; t[7] = 4.83, p = 0.0019) but no significant change in nondeprived eye visual responses (n = 8; green). n.s. denotes not significant. (Lower graph) The same data shown in a plotted as a percentage of baseline amplitude. Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 4 The Number of Transplanted PV+ Cells Does Not Predict the Extent of Plasticity (A) An example coronal section from a MGE recipient. Transplanted PV+ cells (red) are observed in all layers of binocular visual cortex (CTX I-VI) but do not cross the corpus callosum (CC). (B) An example coronal section from a LGE recipient; few transplanted PV+ cells were found in LGE recipients. (C) Quantification of transplanted PV+ cells found in a subset of MGE 35 DAT (magenta), MGE 70 DAT (cyan), and LGE (black) recipients plotted against ocular dominance shifts for each animal. Square points correspond to individuals presented in (A) and (B). No relationship between ocular dominance shift and cell count was observed for MGE recipients (R2 = 0.042; p = 0.46, n.s.). (D) Transplanted PV+ cell spread aligned to the peak PV+ cell count position (0 on the x axis) and averaged for each group to illustrate distribution of transplanted cells. Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 5 Transplantation Reverses Cortical Effects of Visual Deprivation during the Critical Period (A) Timeline of the experimental protocol. Only responses to contralateral eye stimulation were recorded from each hemisphere. (B) Retinotopic maps for an example animal. From left to right: deprived eye 0 DAT (black) and 48 DAT (magenta), and nondeprived eye (green). Scale bar denotes visual field elevation. (C) For animal from (B), cortical responses to visual stimuli across a range of spatial frequencies presented to the deprived eye at 0 DAT (black) and 48 DAT (magenta), and responses from the control (nondeprived) eye (green dashed line) (D) Average deprived eye responses at 15 DAT (black, n = 6) compared to 55 DAT (magenta, n = 5); Lines differ significantly: ANCOVA, p = 0.01. Control eye responses for this group shown as green dashed line (E) Average deprived eye responses at 15 DAT (black, n = 3) and 55 DAT (dashed magenta, n = 3) in dead MGE transplant recipients; lines do not differ significantly. Control eye responses for this group are shown as green dashed line. (F) Average acuity of visual responses to stimulation of deprived and nondeprived eyes in live and dead MGE recipients as well as untreated animals differed across groups H = 28.35, p < 0.0001. Compared to nondeprived control eye (green, n = 18) acuity, deprived eye acuity of live MGE (black solid, n = 7) and dead MGE (black striped, n = 4) recipients at 15 DAT was poor (p < 0.002 and p < 0.05, respectively). By 55 DAT, deprived eye acuity recovered to nondeprived eye levels in live MGE (magenta, n = 5), but not dead MGE (magenta, stripes; n = 3; p < 0.05) recipients. No spontaneous recovery was observed in untreated impaired animals (gray, n = 4, p < 0.03). Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions

Figure 6 Transplantation Reverses Perceptual Deficits in Visually Impaired Animals (A) Schematic of the visual water task. (B) (Top) Representative performance on a visual task for (top) a mouse using the nondeprived eye (green), acuity threshold = 0.46 cpd (gradings on x axis schematically represent the spatial frequency of the visual stimulus); (middle) an untreated mouse using the deprived eye, acuity threshold = 0.30 cpd. Inset graph compares the maximum performance by untreated animals through deprived (black) and nondeprived (green) eyes versus normally sighted animals (gray). (Bottom) A MGE recipient mouse using the deprived eye for the visual task, acuity threshold = 0.51 cpd. (C) Quantification of perceptual acuity across groups (F(3, 37) = 7.52, p = 0.0005). MGE recipients tested using the deprived eye (n = 5, magenta) had acuity equivalent to that of deprived animals using the control (nondeprived) eye (n = 18, green) and to that of naive animals (n = 9, gray). Untreated, impaired animals tested using the deprived eye (n = 9, black) had significantly depressed acuity thresholds compared to MGE recipients (p = 0.007), nondeprived eyes (p = 0.0008), and naive animals (p = 0.003). Error is reported as SEM. Neuron 2015 86, 1055-1066DOI: (10.1016/j.neuron.2015.03.062) Copyright © 2015 Elsevier Inc. Terms and Conditions